Team and data from on-going. studies. This RFA response represents a collaborative effort among scientists from several institutions, including the University of Washington, Vanderbilt University, and Group Health Cooperative (GHC). Nested within the GHC population are several on-going population-based case-control studies designed to assess drug-gene interactions on the outcomes of myocardial infarction (MI), stroke, and new-onset atrial fibrillation (AF). The 3 case groups share a single population-based control group. GHC's computerized pharmacy database provides complete information on all prescription medications. Data collection is funded through 2004, by which time 1053 MI cases, 565 stroke cases, 800 AF cases, and 3249 controls will have been recruited. With resources largely devoted to recruitment, the typical grant included only 1 single nucleotide polymorphism (SNP) per gene in each of a few candidate genes-an approach that ignores the complexity of the underlying linkage disequilibrium or haplotype structure of candidate genes.
Aims. The primary aim of this study is to assess interactions between selected cardiovascular medications and the major candidate-gene variants or haplotypes on the incidence of MI, stroke and AF. The candidate-gene sets-selected on the basis of biology, pharmacology, and information from genome-wide scans-include: (1) 10 genes in the renin-angiotensin system; (2) 10 genes involved in renal sodium transport; (3) 8 genes encoding alpha and beta adrenergic receptors; (4) 8 other genes, including G-proteins, estrogen receptors, and the alpha-1C subunit of the L-type calcium channel. The products of these genes represent the sites of action for many cardiovascular drugs: angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, all classes of diuretics, alpha blockers, beta-blockers, central alpha agonists, calcium antagonists, and estrogens. Both hypothesis-testing and hypothesis-generating analyses are planned. For selected SNPs studied in vitro or in small clinical populations, several specific drug-gene interactions are plausible, and these associations will be evaluated in an hypothesis-testing manner. This component of the study has good to excellent power. The use of single SNPs often fail to capture important components of genetic variation, and the hypothesis-generating part of the study relies on SNP discovery. Already planned for 20 of the 36 genes, SNP discovery will be done in another 16 genes as part of this project. Genotyping an average of 6 to 7 SNPs in each of these 36 genes will allow us to identify the common haplotypes. With the aid new analytic methods such as logic regression, we plan to use these SNP to and haplotype data to evaluate drug-gene interactions on the incidence of MI, stroke, and AF in a systematic fashion. Secondary aims include plans to collect whole-blood mRNA, which will be used to assess the relative expression of different alleles within the same person. While SNP discovery for haplotypes redresses the limited genetic information available from single-SNP studies, multiples testing remains a limitation. The primary defense is replication: we would therefore welcome collaboration from others in the Pharmacogenetics Network in validation efforts.
Showing the most recent 10 out of 68 publications
